当前位置: X-MOL 学术Neural Netw. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Multi-level landmark-guided deep network for face super-resolution
Neural Networks ( IF 6.0 ) Pub Date : 2022-05-05 , DOI: 10.1016/j.neunet.2022.04.026
Cheng Zhuang 1 , Minqi Li 1 , Kaibing Zhang 2 , Zheng Li 1 , Jian Lu 1
Affiliation  

Recent years deep learning-based methods incorporating facial prior knowledge for face super-resolution (FSR) are advancing and have gained impressive performance. However, some important priors such as facial landmarks are not fully exploited in existing methods, leading to noticeable artifacts in the resultant SR face images especially under large magnification. In this paper, we propose a novel multi-level landmark-guided deep network (MLGDN) for FSR. More specifically, to fully exploit the dependencies between low and high resolution images and to reduce network parameters as well as capture more reliable feature representation, we introduce a recursive back-projection network with a particular feedback mechanism for coarse-to-fine FSR. Furthermore, we incorporate an attention fusion module in the front of backbone network to strengthen face components and a feature modulation module to refine features in the middle of backbone network. By this way, the facial landmarks extracted from face images can be fully shared by the modules in different levels, which benefit to produce more faithful facial details. Both quantitative and qualitative performance evaluations on two benchmark databases demonstrate that the proposed MLGDN can achieve more impressive SR results than other state-of-the-art competitors. Code will be available at https://github.com/zhuangcheng31/MLG_Face.git/



中文翻译:

用于人脸超分辨率的多级地标引导深度网络

近年来,基于深度学习的方法结合了面部超分辨率(FSR)的面部先验知识正在取得进步并取得了令人印象深刻的性能。然而,一些重要的先验(例如面部标志)在现有方法中没有得到充分利用,导致生成的 SR 面部图像中出现明显的伪影,尤其是在大放大倍率下。在本文中,我们提出了一种用于 FSR 的新型多级地标引导深度网络 (MLGDN)。更具体地说,为了充分利用低分辨率和高分辨率图像之间的依赖关系并减少网络参数以及捕获更可靠的特征表示,我们引入了具有特定反馈机制的递归反投影网络,用于从粗到细的 FSR。此外,我们在骨干网络前面加入了一个注意力融合模块来加强人脸组件,并在骨干网络中间加入了一个特征调制模块来细化特征。这样,从人脸图像中提取的人脸特征点可以被不同层次的模块充分共享,有利于产生更真实的人脸细节。对两个基准数据库的定量和定性性能评估都表明,与其他最先进的竞争对手相比,提议的 MLGDN 可以获得更令人印象深刻的 SR 结果。代码将在 https://github.com/zhuangcheng31/MLG_Face.git/ 这有利于产生更忠实的面部细节。对两个基准数据库的定量和定性性能评估都表明,与其他最先进的竞争对手相比,提议的 MLGDN 可以获得更令人印象深刻的 SR 结果。代码将在 https://github.com/zhuangcheng31/MLG_Face.git/ 这有利于产生更忠实的面部细节。对两个基准数据库的定量和定性性能评估都表明,与其他最先进的竞争对手相比,提议的 MLGDN 可以获得更令人印象深刻的 SR 结果。代码将在 https://github.com/zhuangcheng31/MLG_Face.git/

更新日期:2022-05-05
down
wechat
bug